AI video analytics for retail stores, using the cameras you already have
You run one store or many, and your cameras record the shop floor all day. Eye AI turns that footage into numbers and proof: how many people came in and when, who entered the stockroom area, and a searchable record of what happened. It works with your existing cameras, and your video is processed on a computer in the store.
What goes wrong on sites like yours
You know takings, but not how many people walked in, or when the busy hours really are.
Stockroom and back-of-house doors are watched only if someone remembers to check the footage.
After an incident, finding the right moment means scrubbing hours of recording.
What Eye AI does here
| Analytic | What it does on your site | Status |
|---|---|---|
| Line crossing and counting | Counts people through each entrance, by direction, hour and day | Live |
| Restricted-area alerts | Logs entry into stockroom and staff-only areas, with a picture | Live |
| Event history and reports | Searchable history of every event, with CSV export for managers | Live |
| Loitering detection | Flags someone staying unusually long at an entrance or an aisle end | Beta |
| Group and crowd clusters | Flags clusters of four or more people near a till or a door | Beta |
| Blocked exits and fire routes | Checks that exits and escape routes stay clear | Launching with pilot partners |
A day on your kind of site
The doors open and Eye AI starts counting entries by the hour. Mid-morning a delivery goes through the back door and the stockroom zone logs it with a picture. Late afternoon the entrance count peaks, and the manager sees the pattern on the day-by-hour grid. After closing, the manager exports the day to a spreadsheet for the weekly review.
Reports and evidence you get
- Footfall by entrance, hour and day
- Back-of-house entry log with pictures
- A weekly CSV for managers and head office
Why existing cameras matter here
Retail already has cameras at the doors and in the back. Using them means you start with the views you have, and your video stays in the store.
Honest limits for this industry
- A busy door with people crossing side by side can merge counts; we check each entrance view during the survey.
- Groups are flagged as clusters, not as a queue length or a wait time.
- Colours and clothing are estimates.
Questions
Can it count people at the door?
Yes. You mark a counting line at each entrance and Eye AI counts people in and out, by hour.
Does it measure queue wait times?
No. It can flag a cluster of people (beta), but it does not report queue length or wait time.
Do you recognise shoppers?
No. Eye AI counts and logs events. It does not identify people.
Does it work across several stores?
Each store has its own dashboard today. A combined multi-store view is on our roadmap.
Which of this looks like your site?
We read every request and reply within 24–48 hours (working days, Monday to Friday) with questions and a proposed solution.
Request a solution